|
|
| Fig. 1: Plutonium-238 pellet radiating from its own heat (Source: Wikimedia Commons) |
An atomic battery is a device that converts nuclear energy into electricity, where the nuclear energy is in the form of nuclear radiation produced by a radioactive isotope. Compared to electrochemical batteries, atomic batteries have a significantly longer life and larger specific power, measured in units watts per kg-fuel, and so they are typically used as sources of power for applications that require uninterrupted power for years. These qualities make atomic batteries useful for space applications in which a constant supply of power is needed over the course of decades, especially in harsh environments or deep space where solar panels are not deployable. Historically, atomic batteries have been used for many deep-space probes to supply a steady electrical power to spacecraft instruments, heaters, and communications for decades at a time; the Mars Curiosity rover is still being powered by the atomic battery that it launched with in 2012. [1] Despite the many times atomic batteries have been deployed in space, they have never been used to for the purpose of powering an electric propulsion (EP) system. The following sections discuss why this is the case and evaluate the possibility of doing so.
Before discussing the application of powering an EP system, we first discuss how atomic batteries work. Atomic batteries house a nuclear fuel which naturally undergoes radioactive decay. The decay process releases a specific amount of energy for each decay event that takes place, in the form of nuclear radiation. Consider an atomic battery that uses a plutonium-238 (Pu-238) fuel, as shown in Fig. 1. Pu-238 emits alpha-radiation at a rate of about 635-trillion decay events per second per kg-fuel and each alpha-particle has an energy of about 5.6 MeV, which means Pu-238 fuel produces a specific-power of about 570 W/kg. [2] The total amount of power supplied by an initial mass of radioactive fuel reduces over time. This is due to the fact that each decay event reduces the total mass of fuel by the mass of an alpha-particle, in turn reducing the total number of decay events occurring at any given moment. The mass of a radioactive fuel pellet over time follows an exponential decay law, where the half-life is a standard measure for indicating the "lifetime" of a radioactive fuel. Pu-238 has a half life of about 88 years, which means that in 88 years a Pu-238 fuel pellet would be half of its initial mass, and thus produce half the total power that it initially did. [2] For example, a 1 kg mass of Pu-238 would initially produce a total power of 570 W, but in 88 years would produce about 285 W of total power. The exponential decay law causes the power output of atomic batteries to be very predictable over its lifetime, and Pu-238's long half-life makes power output predictable for decades of atomic battery operation.
The large specific-power output and long half-life makes Pu-238 a good fuel for atomic batteries, with it being the most commonly used fuel for a specific type of atomic battery called a radioisotope thermoelectric generator (RTG). [2] RTGs are a type of thermoelectric atomic battery in the sense that it converts the roughly 570 W/kg-fuel thermal energy, generated from the radiated alpha-particles colliding with the fuel itself, into electrical energy. The way RTGs convert the thermal energy into electrical energy is via a thermoelectric conversion system, typically with the main thermoelectric convertor being thermocouples. The thermocouples convert thermal power to electrical power by generating a voltage that is proportional to the temperature difference between the Pu-238 fuel and a heat sink due to the Seebeck effect. [2] Even though the Pu-238 fuel has a large thermal specific-power of about 570 W/kg-fuel, the electrical power output of an RTG is typically only on the order of about 25 to 35 watts-electric (We) per kg-fuel, depending on the specific design. [3] This low electrical specific-power is due to the fact that thermocouples inherently have poor thermoelectric conversion efficiency, typically about 5%. [4]
To improve the electrical power output, RTGs with alternative thermoelectric conversion schemes have been developed, with the Stirling radioisotope generator (SRG) being one of the most notable. The SRG is essentially the same as a RTG except that it uses a Stirling engine and alternator combination for thermoelectric conversion instead of using thermocouples. The Stirling engine component converts the generated thermal energy into mechanical energy, and an alternator converts this mechanical energy into electrical energy. This Stirling-based thermoelectric conversion system results in a much higher thermoelectric conversion efficiency of around 25% to 38% and, in turn, a much larger electrical power output of around 120 We/kg-fuel. [5]
 |
| Fig. 2: One of the RTGs aboard NASA's Voyager probes (Source: Wikimedia Commons) |
RTGs have been reliably fielded in space since the 1960s; an example of one the RTGs fielded on the National Aeronautics and Space Administration (NASA) Voyager probes can be seen in Fig. 2. The RTGs used for United States (US) missions are generally composed of modular general-purpose heat source (GPHS) units, which each initially house about 430g of Pu-238 fuel to generate heat, and a thermoelectric convertor, which can be a standard thermocouple system or a Stirling-engine convertor. [6]
The multi-mission-RTG (MMRTG), which is heated by eight GPHSs and uses a lead- tellurium (PbTe) tellurium-silver-germanium-antimony (TAGS) (PbTe/TAGS) thermocouple-based thermoelectric convertor. The MMRTG provides an initial power of about 110 We and total mass of 40 kg; an initial system-specific power of 2.8 We/kg. It is designed to be used for a minimum lifetime of 14 years and capable of reliable operation in planetary atmospheres, which is why the MMRTG was fielded on the Mars Curiosity rover as its power source. [7]
The GPHS-RTG has been used for missions that require a larger total power output than the MMRTG, and is composed of sixteen GPHS units heating a Si-Ge thermocouple-based thermoelectric convertor. The GPHS-RTG generates an initial power of about 245 We and total mass of 56 kg; an initial system-specific power of about 4.4 We/kg. [8] The GPHS-RTG has been fielded on several NASA missions, including most recently the 2015 flyby of Pluto in the New Horizons mission. [8]
The SRG could be used for missions that require a total power output larger than both the MMRTG and GPHS-RTG. Although SRGs have been developed and tested, they have never been fielded in space due to their lower technology readiness level (TRL). The Stirling Research Lab at NASA Glenn Research Center (GRC), for example, has been working on SRG and String-based thermoelectric converter design. GRC's most recent SRG is based on a Stirling thermoelectric convertor design called the Sunpower Robust Stirling Convertor (SRSC). [5] NASA GRC collaborated with the US Department of Energy (DOE) and Aerojet Rocketdyne to design an SRG based on six GPHS units and eight SRSCs, with a total initial power output of about 350 We and total mass of 112 kg; an initial system-specific power of about 3.1 We/kg. [5,9]
The specifications of the MMRTG, GPHS-RTG, and SRG are summarized in Table 1:
|
||||||||||||||||||||||||||||
| Table 1: List of specifications for three types of RTGs. [5-9] |
Despite the many times RTGs have been fielded in space, they have never been used for the purpose of powering an EP system. An RTG-powered EP system is termed radioisotope electric propulsion (REP) and has been sparsely discussed in the scientific literature. A general consensus on REP is that a standalone REP system has a potential use case for outer solar system (Saturn to Pluto) missions that include the following criteria: [10,11]
The EP system has a sub-kW power consumption
The spacecraft mass (excluding the RTGs and EP system) is less than 300 kg
The mission takes place where solar panels are not usable
The first and second criteria are necessary for an REP system because of the inherently large power-specific mass of RTG sources. For example, the RTG units described in Table 1 have 100s of kg/kWe power-specific masses, meaning for every kilowatt of electrical power that the EP system must consume, 100s of kilograms must be added to the spacecraft. This causes the EP power consumption and spacecraft mass to both be constrained to sub-kW and sub-300 kg, respectively, or else the mission would take an impractical amount of time. [10]
The third criterion, that the mission takes place where solar panels are not usable, is because if solar panels were usable for the mission then they would almost always be the more appropriate choice for a power source since solar panel system-specific power can exceed 100 We/kg, much larger than RTG system-specific power of 2 to 5 We/kg. [10,12]
When considering missions to the outer solar system powered by REP, an REP system of specific mass less than 150 kg/kWe is needed to complete the mission in a reasonable amount of time. [10] Consider an REP system that consists of an RTG power source and a sub-kW EP system with mass of 20 kg. Based on the sub-kW power constraint, a minimum REP system-specific mass (RTG plus EP system) for each type of RTG power source is calculated and shown in Table 2:
|
||||||||||||||||||||
| Table 2: List of specific masses of an REP system for three different types of RTG power sources. The REP system considered is comprised of an RTG power source and an arbitrary 20 kg sub-kW EP system. [5-9] |
Based on the results of Table 2, none of the discussed RTG types are currently a viable power source for an REP system based on the 150 kg/kWe specific mass requirement for an outer solar system mission. However, considering the SRG has a large thermoelectric conversion efficiency and is also a less-established RTG technology compared to the MMRTG and GPHS-RTG, it may be worth trying to reduce the total mass of the SRG unit in order to improve the SRG's specific power output. For example, if the mass of the SRG unit was hypothetically reduced from 112 kg to 42 kg and the produced electrical power remained 350 We, this results in a dramatic specific power increase to 8.3 We/kg; in turn, two SRG units and a 20 kg sub-kW EP system would have a REP system- specific mass of 149 kg/kWe, potentially enabling an REP outer solar system mission. This is an ambitious goal, however, because the mass of the SRG would need to be reduced to about 38% of what it currently is. To determine whether the same RTGs could be used in an REP system for other mission types requires further considerations on specific mission requirements.
REP could potentially enable near-future outer solar system missions fielding a sub-kW EP system aboard a sub-300 kg spacecraft when solar panels are not a viable power source. However, the system-specific mass of currently-available RTGs when considering MMRTG, GPHS-RTG, and SRG types is too large to enable missions that require a sub-150 kg/kWe REP system. Thus, specific power of current RTG technology must be improved before RTGs can enable an outer solar system mission. Considering SRGs have an inherently more efficient thermoelectric convertor and are less established than GPHS-RTGs and MMRTGs, it may be worth the efforts of improving SRG system-specific power for REP application; although this determination would require further evaluation of state-of-the-art SRG design. To evaluate REP viability for other mission types requires further consideration on specific mission requirements.
© Adam Tuckey. The author warrants that the work is the author's own and that Stanford University provided no input other than typesetting and referencing guidelines. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.
[1] J. Belanger, "Powering NASA's Curiosity," Physics 240, Stanford University, Fall 2012.
[2] P. Dedeler, "Radioisotope Power for Fueling Space Missions," Physics 240, Stanford University, Fall 2023.
[3] G. Bennett, "Space Nuclear Power: Opening the Final Frontier," American Institute of Astronautics and Aeronautics, "AIAA 2006-4191, 4th Intl. Energy Conversion Engineering Conference and Exhibit, 26 Jun 06.
[4] H. J. Goldsmid, The Physics of Thermoelectric Energy Conversion (IOP Concise Physics, 2017).
[5] S. Wilson, "Development of Stirling Convertors for Radioisotope and Fission Power Systems ," in Nuclear and Emerging Technologies for Space (American Nuclear Society, 2023), p. 297.
[6] G. I. Bennett et al., "The General-Purpose Heat Source Radioisotope Thermoelectric Generator: A Truly General-Purpose Space RTG," AIP Conf. Proc. 969, 663 (2008).
[7] "Multi-Mission Radioisotope Thermoelectric Generator (MMRTG)," U.S. National Aeronautics and Space Administration, May 2020.
[8] "Next Generation Radioisotope Thermoelectric Generators," U.S. National Aeronautics and Space Administration, Novedmber 2022.
[9] E. Soendker and A. Poehls, "Controller Development for the Dynamic Radioisotope Power System," IEEE 10521135, 2024 IEEE Aerospace Conference, 2 Mar 24.
[10] S. Oleson et al., "Radioisotope Electric Propulsion for Fast Outer Planetary Orbiters," U.S. National Aeronautics and Space Administration, NASA/TM-2002-211893, July 2002.
[11] G. R. Schmidt et al., "Radioisotope Electric Propulsion (REP): A Near-Term Approach to Nuclear Propulsion," Acta Astronaut. 66, 501 (2010).
[12] "ROSA (Roll-Out Solar Array)," Redwire Space, October 2025.
.jpg){kind=link}